Add a late optimization to BranchFolding that hoist common instruction sequences

at the start of basic blocks to their common predecessor. It's actually quite
common (e.g. about 50 times in JM/lencod) and has shown to be a nice code size
benefit. e.g.

        pushq   %rax
        testl   %edi, %edi
        jne     LBB0_2
## BB#1:
        xorb    %al, %al
        popq    %rdx
        ret
LBB0_2:
        xorb    %al, %al
        callq   _foo
        popq    %rdx
        ret

=>

        pushq   %rax
        xorb    %al, %al
        testl   %edi, %edi
        je      LBB0_2
## BB#1:
        callq   _foo
LBB0_2:
        popq    %rdx
        ret

rdar://9145558


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@131172 91177308-0d34-0410-b5e6-96231b3b80d8
This commit is contained in:
Evan Cheng
2011-05-11 01:03:01 +00:00
parent b38fe4b52d
commit 7e20a574eb
4 changed files with 296 additions and 10 deletions

View File

@ -265,7 +265,7 @@ bool IfConverter::runOnMachineFunction(MachineFunction &MF) {
if (!TII) return false;
// Tail merge tend to expose more if-conversion opportunities.
BranchFolder BF(true);
BranchFolder BF(true, false);
bool BFChange = BF.OptimizeFunction(MF, TII,
MF.getTarget().getRegisterInfo(),
getAnalysisIfAvailable<MachineModuleInfo>());
@ -399,7 +399,7 @@ bool IfConverter::runOnMachineFunction(MachineFunction &MF) {
BBAnalysis.clear();
if (MadeChange && IfCvtBranchFold) {
BranchFolder BF(false);
BranchFolder BF(false, false);
BF.OptimizeFunction(MF, TII,
MF.getTarget().getRegisterInfo(),
getAnalysisIfAvailable<MachineModuleInfo>());